No belief propagation required: Belief space planning in high-dimensional state spaces via factor graphs, the matrix determinant lemma, and re-use of calculation.
Dmitry KopitkovVadim IndelmanPublished in: Int. J. Robotics Res. (2017)
Keyphrases
- belief space
- factor graphs
- belief propagation
- state space
- graphical models
- high dimensional
- message passing
- belief state
- approximate inference
- stereo matching
- classical planning
- initial state
- markov random field
- graph cuts
- planning problems
- partially observable
- heuristic search
- dynamic programming
- motion planning
- reinforcement learning
- markov networks
- dynamic bayesian networks
- pairwise
- exact inference
- optimal policy
- probabilistic model
- latent variables
- probabilistic inference
- markov chain
- state variables
- random variables
- probabilistic graphical models
- particle filter
- search space
- bayesian networks
- dynamical systems
- energy function
- partially observable markov decision processes
- feature space
- planning graph
- image retrieval
- belief networks
- expectation maximization